The MCS-4 as it was first called, could only be sold to Busicom who had paid for most of the development. A little later on a slump in the calculator market forced Busicom to ask Intel to lower the price of the MCS-4. Intel agreed but only if they could sell the chip set for non-calculator applications - they agreed.
Intel’s marketing department weren’t so sure that this was a good thing. The sales of minicomputers were so low that volume production of the MCS-4 looked unlikely. Of course, as we know to day, the sales of minicomputers had no bearing on the potential sales of microcomputers, but the MCS-4 was seen as a minicomputer replacement.
The problem was solved by Arthur Rock, Intel's first investor, who simply recognised a good thing when he saw it and told the board of directors so. The MCS-4 was announced at the end of 1971 and they sold a modest $85,000 worth of them by the start of 1972.
Even with the 4004, Intel thought it had a big computer to sell!
The really important thing about the 4004 from the electronic engineer's point of view was that it was a family of chips that were easy to use together. You took a handful of chips all called 400-something and you put them together to make your system with as much I/O and memory as you could afford. Programmers on the other hand were less than impressed by the ad-hoc archecture and instruction set. Later programmers would get better options, but by then the history had already been written that the 4004 would become the starting point for all of the processors we use today.
Intel's nostalgic look back at the 4004 on its 40th aniversary.
Beyond the 4004
It is too simple to look back and see the history of Intel processors as a steady increase in the serial numbers used to identify them.
While Intel was working on the MCS-4 another project was going in the direction of a microprocessor implementation. CTC wanted a chip set for a VDU and, because this had to work with characters, the processor that Ted Hoff decided to design had to be an 8-bit processor. CTC, however had other ideas. They took their plans to Texas Instruments who designed an 8-bit processor for them. It was never used because CTC decided to implement their design using TTL logic chips by the bucket load.
Although the CTC project came to a sudden end, Hal Feeney continued with the design assisted by Stan Mazor and Ted Hoff.
The first 8-bit processor, the 8008
The 8008, as it was designated, was introduced in 1972 as the first general purpose 8-bit microprocessor. It supported 45 instructions, a 30us execution time, six general purpose registers, an address space of 16Kbytes and it was packaged in an 18-pin DIP. If you put it together with around 20 other chips you had a machine that was recognisable as a computer.
The 8008 rather than the 4004 started the microprocessor revolution and led directly to the 8080 and eventually the 80486, the Pentium and beyond.
Intel had its second big hit and sales reach $66 million by 1978. Noyce was worth $18.5 million and the company’s stock value had tripled.
Robert Noyce, Gordon Moore and Andy Grove with the 8080 design
In this third and final part of our look at the early days of the microcomputer revolution as it happened in the UK, we focus on the amazing and idiosyncratic SWTP family of computers which played the [ ... ]
The Harvard Mark I now has an established place in the history of computing. However, without financial help from IBM it would never have materialized. It is more than 75 years since the memorandum fo [ ... ]